Cellulose esters are well known compounds (“Cellulose Derivatives”, Ben P. Rouse, Kirk-Othmer Encyclopedia of Chemical Technology, vol 4, 1964, 616-683). The most common cellulose esters are comprised of aliphatic C2-C4 substitutents. Examples of such cellulose esters include cellulose acetate (CA), cellulose propionate (CP), cellulose butyrate (CB), cellulose acetate propionate (CAP), and cellulose acetate butyrate (CAB). Examples of the utility of such cellulose esters can be found in Prog. Polym. Sci. 2001, 26, 1605-1688. Cellulose esters are generally prepared by first converting cellulose to a cellulose triester before hydrolyzing the cellulose triester in an acidic aqueous media to the desired degree of substitution (DS, the number of substitutents per anhydroglucose monomer). Aqueous acid catalyzed hydrolysis of cellulose triacetate yields a random copolymer that can consist of 8 different monomers depending upon the final DS (Macromolecules 1991, 24, 3050).
Processes that provide non-random copolymers via hydrolysis of cellulose triesters are known. Direct esterification to less than fully substituted cellulose esters is also known. Depending upon the precise reaction conditions, it is possible to obtain a non-random cellulose ester by this type of process.
Recently, there have been accounts of attempts at the preparation of regioselectively substituted cellulose derivatives. For the purposes of this invention, regioselective substitution means the exclusive or preferential placement or removal of a substituent at the C2, C3, or C6 hydroxyls of the anhydroglucose monomer of cellulose. Controlled placement of the substituent can lead to a homopolymer or a copolymer of cellulose with specific monomer content. That is, a cellulose derivative with a specific substitution pattern within the anhydroglucose monomer and a controlled sequence along the cellulose polymer chain is obtained.
Prior methods leading to the formation of regioselective substituted cellulose derivatives rely on the use of temporary protecting groups and either requires the use of cellulose solvents so that the protecting group can be installed in a homogeneous reaction mixture or mercerized cellulose that has sufficient reactivity.
The preparation of certain formate esters of carbohydrates and polysaccharides is known. The isolated cellulose formates are typically used as unstable intermediates for subsequent reactions due to reported instability of the formate ester and reactivity toward other functional groups. As a result, the formation of mixed cellulose formate derivatives has received little attention, and few reports of the formation of a mixed cellulose formate esters exists. GB 568,439 (1945) teaches the preparation of cellulose acetate formate which is produced by mixing cellulose with an acetic and formic mixed anhydride in the presence of a catalyst.
Only a few classes of carboxylated cellulose esters are known. One example of this class of cellulose ester derivatives is carboxymethyl cellulose esters described for example in U.S. Pat. Nos. 5,668,273; 5,792,856; and 5,994,530. These cellulose derivatives are cellulose ether esters in which an intervening ether linkage attaches a carboxylate to the anhydroglucose units of the cellulose chain. These derivatives are formed by esterifying carboxymethyl cellulose (an ether) to the fully substituted carboxymethyl cellulose ester followed by hydrolysis to the desired ester DS. This class of carboxylated cellulose esters offers the advantage of a non-hydrolysable carboxylate linkage. The disadvantage is that the method of preparation is a two-step process requiring the preparation and isolation of the carboxymethyl cellulose prior to esterification. Furthermore, one cannot obtain a consistent, homogeneous distribution of carboxymethyl substitutents along the cellulose backbone.
Another class of carboxylated cellulose esters is those in which the carboxylate functionality is attached to the cellulose backbone via an ester linkage. An example of this class is cellulose acetate phthalate and the like which are described in U.S. Pat. No. 3,489,743. In general, these cellulose ester derivatives are formed by first preparing a neutral, randomly substituted cellulose ester, e.g. a CA, with the desired DS. In a second reaction, the carboxylate functionality is installed by treating the cellulose ester with an anhydride such as phthalic anhydride.
An additional class of carboxylated cellulose esters is those, which result from ozonolysis of cellulose esters in the solid state (Sand, I. D., Polymer Material Science Engineering, 1987, 57-63; U.S. Pat. No. 4,590,265). Ozonolysis of cellulose ester provides a polymer that contains not only carboxylates but also aldehydes, ketone, and peroxides as well. The process results in significant loss in polymer molecular weight and relatively low levels of oxidation. Furthermore, the process is not specific in that any of the cellulose ester hydroxyls can be oxidized.
Oxidation of carbohydrates and polysaccharides is a very important process in the chemical industry and a number of useful catalysts for this transformation have been developed. Some of the most useful catalysts belong to the class of compounds referred to as nitroxyl or nitroxide radicals. Typically, these compounds are secondary amine nitroxides with the general structure shown below.

Of the secondary amine N-oxides, the cyclic hindered nitroxyls belonging to the piperidine series have proven to be the most interesting. There are many routes for the synthesis of cyclic nitroxyl derivatives in the piperidine series. The vast majority of the methods use 4-oxo-2,2,6,6-tetramethylpiperdine (triacetoneamine) as the common starting material, which is generally prepared by the cyclocondensation of acetone and ammonia (Sosnovsky, G.; Konieczny, M., Synthesis, 1976, 735-736). Triacetoneamine serves as a common intermediate for the synthesis of a number of different derivatives such as those shown in Scheme 1 below. Of the derivatives shown in Scheme 1, 2,2,6,6-tetramethylpiperdine-N-oxyl (5, TEMPO) has proven to be the cyclic nitroxyl used in most studies involving oxidation of alcohols.

Oxidation of alcohols with TEMPO under acidic conditions converts primary and secondary alcohols to aldehydes and ketones, respectively (Bobbit, J. M.; Ma, Z., J. Org. Chem. 1991, 56, 6110-6114). Generally, over oxidation is not observed, but two molar equivalents of TEMPO per mole of substrate are required for the oxidation of the alcohol. That is, the reaction is not catalytic.
The use of stoichiometric amounts of TEMPO or its analogues for oxidation of alcohols can be expensive and create difficulties in isolation of the product. As a result, work in this area has focused on catalytic processes that regenerate the nitrosonium ion in situ by the use of primary and/or terminal oxidants. The primary oxidant oxidizes the hydroxy amine back to the nitrosonium ion, and the terminal oxidant serves to regenerate the primary oxidant. In some cases, the primary oxidant functions as both the primary and terminal oxidant.
It is possible to oxidize alcohols under acidic conditions using catalytic amounts of TEMPO or its analogues. However, the solvents for this process are limited and acid sensitive substrates typically are destroyed under these conditions. Furthermore, primary and secondary alcohols are typically converted to aldehydes and ketones, respectively, rather than to a carboxylic acid.
TEMPO catalyzed oxidations of primary alcohols in nonaqueous reaction media under acidic conditions (pH<4) can give almost exclusively the corresponding aldehyde. In aqueous media, some subsequent conversion of the aldehyde to a carboxylate is observed, but the aldehyde to carboxylic acid ratio remains high. As a result, oxidation of primary alcohols of polysaccharides and carbohydrates under acidic conditions using prior art TEMPO catalyzed conditions is of limited utility due to the fact that the extent of oxidation is limited and the reaction media is not suitable for many substrates.
As a result, research concerning the oxidation of primary alcohols of polysaccharides and carbohydrates with TEMPO and TEMPO analogues has focused on oxidation under alkaline conditions. Because most polysaccharides and carbohydrates have limited solubility in organic solvents, most investigations have focused on the use of an aqueous reaction media.
Typical pH and temperature for TEMPO catalyzed oxidation of polysaccharides, such as starch, are in the range of 8.5-11.5 at a temperature of −10 to 25° C. (Tetrahedron Letters 1999, 40, 1201-1202; Macromolecules 1996, 29, 6541-6547; Tetrahedron 1995, 51, 8023-32; Carbohydr. Polym. 2000, 42, 51-57; Carbohydr. Res. 2000, 327, 455-461; Carbohydr. Res. 1995, 269, 89-98; WO 96/38484; Recl. Trav. Chim. Pays-Bas 1994, 113, 165-6; Carbohydr. Res. 2001, 330, 21-29; Carbohydr. Lett. 1998, 3, 31-38; EP 1077221 A1; Synthesis 1999, 5, 864-872; J. Mol. Catal. A: Chem. 1999, 150, 31-36). In most cases, the primary oxidant is NaBr and the terminal oxidant is NaOCl.
Oxidation of polysaccharides and carbohydrates under alkaline conditions using analogues of TEMPO and other primary oxidants has also been investigated (Carbohydrate Research 2000, 328, 355-363; J. Mol. Catal. A: Chem. 2001, 170, 35-42; J. Catal. 2000, 194, 343-351; Proc. Electrochem. Soc. 1993, 260-7; Carbohydr. Res. 1995, 268, 85-92; EP 0979826 A1; U.S. Pat. No. 5,831,043; U.S. 2001/0034442 A1).
It is difficult, if not impossible, to oxidize cellulose esters using TEMPO under alkaline conditions. One problem is that nearly all cellulose esters are insoluble in water. Additionally, the pH and temperatures commonly employed can lead to rapid and undesirable cleavage of the acyl substitutents. Furthermore, the polymer backbone is rapidly cleaved under these reaction conditions.
Most of the studies involving TEMPO catalyzed oxidation of polysaccharides have involved water-soluble polysaccharides or polysaccharides that are sufficiently reactive so that they can be treated as a suspension in H2O. Attempts to extend TEMPO mediated oxidations to cellulose have met with limited success. Cellulose can be oxidized to a water-soluble polyuronic acid after mercerization in NaOH or after regeneration of the cellulose (Cellulose 2002, 9, 75-81; Cellulose 1998, 5, 153-164).
In view of the previous discussion, it would be useful to have routes to access carboxylated cellulose ester derivatives with high acid numbers wherein the carboxyl group is attached directly to the cellulose backbone by a carbon-carbon bond. Preferably, such a route would be versatile allowing access to carboxylated cellulose esters having a wide range of acid numbers. It would also be desirable that the carboxylates be randomly distributed within the cellulose ester polymer. Through functionalization of the intermediate aldehyde, the corresponding cationic or zwitterionic cellulose ester derivatives could also be accessed.